diff --git a/src/xercesc/util/XMLUri.cpp b/src/xercesc/util/XMLUri.cpp
index 636e04e1be101b060743853b405b08f526a61991..a9df924326f68e9387b3da6be203466ed2bed1e0 100644
--- a/src/xercesc/util/XMLUri.cpp
+++ b/src/xercesc/util/XMLUri.cpp
@@ -1,7 +1,7 @@
/*
* The Apache Software License, Version 1.1
*
- * Copyright (c) 2001 The Apache Software Foundation. All rights
+ * Copyright (c) 2001-2003 The Apache Software Foundation. All rights
* reserved.
*
* Redistribution and use in source and binary forms, with or without
@@ -82,7 +82,8 @@ XERCES_CPP_NAMESPACE_BEGIN
const XMLCh XMLUri::RESERVED_CHARACTERS[] =
{
chSemiColon, chForwardSlash, chQuestion, chColon, chAt,
- chAmpersand, chEqual, chPlus, chDollarSign, chComma, chNull
+ chAmpersand, chEqual, chPlus, chDollarSign, chComma, chOpenSquare,
+ chCloseSquare, chNull
};
//
@@ -113,6 +114,16 @@ const XMLCh XMLUri::USERINFO_CHARACTERS[] =
chDollarSign, chPeriod, chNull
};
+// pchar plus ';' and '/'.
+// pchar = unreserved | escaped |
+// ":" | "@" | "&" | "=" | "+" | "$" | ","
+const XMLCh XMLUri::PATH_CHARACTERS[] =
+{
+ chSemiColon, chForwardSlash, chColon, chAt, chAmpersand,
+ chEqual, chPlus, chDollarSign, chComma, chNull
+};
+
+
// ---------------------------------------------------------------------------
// Local methods and data
// ---------------------------------------------------------------------------
@@ -397,27 +408,31 @@ void XMLUri::initialize(const XMLUri* const baseURI
int index = 0;
bool foundScheme = false;
- // Check for scheme, which must be before `/'. Also handle names with
- // DOS drive letters ('D:'), so 1-character schemes are not allowed.
- int colonIdx = XMLString::indexOf(trimedUriSpec, chColon);
- int slashIdx = XMLString::indexOf(trimedUriSpec, chForwardSlash);
-
- if ((colonIdx < 2) ||
- (colonIdx > slashIdx && slashIdx != -1) )
- {
- int fragmentIdx = XMLString::indexOf(trimedUriSpec, chPound); //'#'
- // A standalone base is a valid URI according to spec
- if ( !baseURI && fragmentIdx != 0 )
+ // Check for scheme, which must be before `/', '?' or '#'.
+ // Also handle names with DOS drive letters ('D:'),
+ // so 1-character schemes are not allowed.
+ int colonIdx = XMLString::indexOf(trimedUriSpec, chColon);
+ int slashIdx = XMLString::indexOf(trimedUriSpec, chForwardSlash);
+ int queryIdx = XMLString::indexOf(trimedUriSpec, chQuestion);
+ int fragmentIdx = XMLString::indexOf(trimedUriSpec, chPound);
+
+ if ((colonIdx < 2) ||
+ (colonIdx > slashIdx && slashIdx != -1) ||
+ (colonIdx > queryIdx && queryIdx != -1) ||
+ (colonIdx > fragmentIdx && fragmentIdx != -1))
{
- ThrowXML(MalformedURLException, XMLExcepts::XMLNUM_URI_No_Scheme);
+ // A standalone base is a valid URI according to spec
+ if ( colonIdx == 0 || (!baseURI && fragmentIdx != 0) )
+ {
+ ThrowXML(MalformedURLException, XMLExcepts::XMLNUM_URI_No_Scheme);
+ }
+ }
+ else
+ {
+ foundScheme = true;
+ initializeScheme(trimedUriSpec);
+ index = XMLString::stringLen(fScheme)+1;
}
- }
- else
- {
- foundScheme = true;
- initializeScheme(trimedUriSpec);
- index = XMLString::stringLen(fScheme)+1;
- }
// It's an error if we stop here
if (index == trimedUriSpecLen || (foundScheme && (trimedUriSpec[index] == chPound)))
@@ -678,12 +693,12 @@ void XMLUri::initializeAuthority(const XMLCh* const uriSpec)
{
int index = 0;
- int start = 0;
+ int start = 0;
const int end = XMLString::stringLen(uriSpec);
//
// server = [ [ userinfo "@" ] hostport ]
- // userinfo is everything up @,
+ // userinfo is everything up @,
//
XMLCh* userinfo = (XMLCh*) fMemoryManager->allocate
(
@@ -696,7 +711,7 @@ void XMLUri::initializeAuthority(const XMLCh* const uriSpec)
{
XMLString::subString(userinfo, &(uriSpec[start]), 0, index);
index++; // skip the @
- start += index;
+ start += index;
}
else
{
@@ -705,25 +720,41 @@ void XMLUri::initializeAuthority(const XMLCh* const uriSpec)
//
// hostport = host [ ":" port ]
- // host is everything up to ':'
+ // host is everything up to ':', or up to
+ // and including ']' if followed by ':'.
//
- XMLCh* host = (XMLCh*) fMemoryManager->allocate
+ XMLCh* host = (XMLCh*) fMemoryManager->allocate
(
(end+1) * sizeof(XMLCh)
);//new XMLCh[end+1];
ArrayJanitor<XMLCh> hostName(host, fMemoryManager);
- index = XMLString::indexOf(&(uriSpec[start]), chColon);
+
+ // Search for port boundary.
+ if (start < end && uriSpec[start] == chOpenSquare)
+ {
+ index = XMLString::indexOf(&(uriSpec[start]), chCloseSquare);
+ if (index != -1)
+ {
+ // skip the ']'
+ index = ((start + index + 1) < end
+ && uriSpec[start + index + 1] == chColon) ? index+1 : -1;
+ }
+ }
+ else
+ {
+ index = XMLString::indexOf(&(uriSpec[start]), chColon);
+ }
- if ( index != -1)
+ if ( index != -1 )
{
XMLString::subString(host, &(uriSpec[start]), 0, index);
index++; // skip the :
- start +=index;
+ start +=index;
}
else
{
XMLString::subString(host, &(uriSpec[start]), 0, end-start);
- start=end;
+ start = end;
}
// port is everything after ":"
@@ -758,7 +789,7 @@ void XMLUri::initializeAuthority(const XMLCh* const uriSpec)
//
setHost(host);
setPort(port);
- setUserInfo(userinfo);
+ setUserInfo(userinfo);
}
// scheme = alpha *( alpha | digit | "+" | "-" | "." )
@@ -785,55 +816,111 @@ void XMLUri::initializeScheme(const XMLCh* const uriSpec)
void XMLUri::initializePath(const XMLCh* const uriSpec)
{
- if ( !uriSpec )
+ if ( !uriSpec )
{
ThrowXML1(MalformedURLException
, XMLExcepts::XMLNUM_URI_Component_Empty
, errMsg_PATH);
- }
+ }
- int index = 0;
- int start = 0;
+ int index = 0;
+ int start = 0;
int end = XMLString::stringLen(uriSpec);
- XMLCh testChar;
+ XMLCh testChar;
- // path - everything up to query string or fragment
- while (index < end)
+ // path - everything up to query string or fragment
+ if (start < end)
{
- testChar = uriSpec[index];
- if (testChar == chQuestion || testChar == chPound)
- {
- break;
- }
-
- // check for valid escape sequence
- if (testChar == chPercent)
+ // RFC 2732 only allows '[' and ']' to appear in the opaque part.
+ if (!getScheme() || uriSpec[start] == chForwardSlash)
{
- if (index+2 >= end ||
- !XMLString::isHex(uriSpec[index+1]) ||
- !XMLString::isHex(uriSpec[index+2]))
+ // Scan path.
+ // abs_path = "/" path_segments
+ // rel_path = rel_segment [ abs_path ]
+ while (index < end)
{
- XMLString::moveChars(value1, &(uriSpec[index]), 3);
- value1[3] = chNull;
- ThrowXML2(MalformedURLException
- , XMLExcepts::XMLNUM_URI_Component_Invalid_EscapeSequence
- , errMsg_PATH
- , value1);
- }
+ testChar = uriSpec[index];
+ if (testChar == chQuestion || testChar == chPound)
+ {
+ break;
+ }
+
+ // check for valid escape sequence
+ if (testChar == chPercent)
+ {
+ if (index+2 >= end ||
+ !XMLString::isHex(uriSpec[index+1]) ||
+ !XMLString::isHex(uriSpec[index+2]))
+ {
+ XMLString::moveChars(value1, &(uriSpec[index]), 3);
+ value1[3] = chNull;
+ ThrowXML2(MalformedURLException
+ , XMLExcepts::XMLNUM_URI_Component_Invalid_EscapeSequence
+ , errMsg_PATH
+ , value1);
+ }
+ }
+ else if (!isUnreservedCharacter(testChar) &&
+ !isPathCharacter(testChar))
+ {
+ value1[0] = testChar;
+ value1[1] = chNull;
+ ThrowXML2(MalformedURLException
+ , XMLExcepts::XMLNUM_URI_Component_Invalid_Char
+ , errMsg_PATH
+ , value1);
+ }
+
+ index++;
+ }//while (index < end)
}
- else if (!isReservedCharacter(testChar) &&
- !isUnreservedCharacter(testChar))
+ else
{
- value1[0] = testChar;
- value1[1] = chNull;
- ThrowXML2(MalformedURLException
- , XMLExcepts::XMLNUM_URI_Component_Invalid_Char
- , errMsg_PATH
- , value1);
- }
+ // Scan opaque part.
+ // opaque_part = uric_no_slash *uric
+ while (index < end)
+ {
+ testChar = uriSpec[index];
+ if (testChar == chQuestion || testChar == chPound)
+ {
+ break;
+ }
- index++;
- }//while (index < end)
+ // check for valid escape sequence
+ if (testChar == chPercent)
+ {
+ if (index+2 >= end ||
+ !XMLString::isHex(uriSpec[index+1]) ||
+ !XMLString::isHex(uriSpec[index+2]))
+ {
+ XMLString::moveChars(value1, &(uriSpec[index]), 3);
+ value1[3] = chNull;
+ ThrowXML2(MalformedURLException
+ , XMLExcepts::XMLNUM_URI_Component_Invalid_EscapeSequence
+ , errMsg_PATH
+ , value1);
+ }
+ }
+ // If the scheme specific part is opaque, it can contain '['
+ // and ']'. uric_no_slash wasn't modified by RFC 2732, which
+ // I've interpreted as an error in the spec, since the
+ // production should be equivalent to (uric - '/'), and uric
+ // contains '[' and ']'.
+ else if (!isUnreservedCharacter(testChar) &&
+ !isReservedCharacter(testChar))
+ {
+ value1[0] = testChar;
+ value1[1] = chNull;
+ ThrowXML2(MalformedURLException
+ , XMLExcepts::XMLNUM_URI_Component_Invalid_Char
+ , errMsg_PATH
+ , value1);
+ }
+
+ index++;
+ }//while (index < end)
+ }
+ } //if (start < end)
if (getPath())
{
@@ -843,8 +930,8 @@ void XMLUri::initializePath(const XMLCh* const uriSpec)
fPath = (XMLCh*) fMemoryManager->allocate((index+1) * sizeof(XMLCh));//new XMLCh[index+1];
XMLString::subString(fPath, uriSpec, start, index);
- // query - starts with ? and up to fragment or end
- if (testChar == chQuestion)
+ // query - starts with ? and up to fragment or end
+ if (testChar == chQuestion)
{
index++;
start = index;
@@ -870,8 +957,8 @@ void XMLUri::initializePath(const XMLCh* const uriSpec)
, value1);
}
}
- else if (!isReservedCharacter(testChar) &&
- !isUnreservedCharacter(testChar))
+ else if (!isUnreservedCharacter(testChar) &&
+ !isReservedCharacter(testChar))
{
value1[0] = testChar;
value1[1] = chNull;
@@ -895,8 +982,8 @@ void XMLUri::initializePath(const XMLCh* const uriSpec)
XMLString::subString(fQueryString, uriSpec, start, index);
}
- // fragment - starts with #
- if (testChar == chPound)
+ // fragment - starts with #
+ if (testChar == chPound)
{
index++;
start = index;
@@ -918,8 +1005,8 @@ void XMLUri::initializePath(const XMLCh* const uriSpec)
, value1);
}
}
- else if (!isReservedCharacter(testChar) &&
- !isUnreservedCharacter(testChar))
+ else if (!isUnreservedCharacter(testChar) &&
+ !isReservedCharacter(testChar))
{
value1[0] = testChar;
value1[1] = chNull;
@@ -1034,8 +1121,7 @@ void XMLUri::setUserInfo(const XMLCh* const newUserInfo)
void XMLUri::setHost(const XMLCh* const newHost)
{
- if ( !newHost ||
- XMLString::isAllWhiteSpace(newHost))
+ if ( !newHost )
{
if (getHost())
fMemoryManager->deallocate(fHost);//delete [] fHost;
@@ -1323,34 +1409,35 @@ bool XMLUri::isURIString(const XMLCh* const uricString)
// domainlabel = alphanum | alphanum *( alphanum | "-" ) alphanum
// toplabel = alpha | alpha *( alphanum | "-" ) alphanum
//
-// IPv4address = 1*digit "." 1*digit "." 1*digit "." 1*digit
+// IPv4address = 1*3DIGIT "." 1*3DIGIT "." 1*3DIGIT "." 1*3DIGIT
//
bool XMLUri::isWellFormedAddress(const XMLCh* const addrString)
{
- if (!addrString)
- return false;
-
- //
- // check length
- //
- XMLCh* tmpAddr = XMLString::replicate(addrString, XMLPlatformUtils::fgMemoryManager);
- ArrayJanitor<XMLCh> janName(tmpAddr, XMLPlatformUtils::fgMemoryManager);
- XMLString::trim(tmpAddr);
- if ((!tmpAddr || !*tmpAddr) ||
- (XMLString::stringLen(tmpAddr) > 255) )
+ // Check that we have a non-zero length string.
+ if (!addrString || !*addrString)
return false;
+
+ // Get address length.
+ int addrStrLen = XMLString::stringLen(addrString);
+
+ // Check if the host is a valid IPv6reference.
+ if (*addrString == chOpenSquare)
+ {
+ return isWellFormedIPv6Reference(addrString, addrStrLen);
+ }
//
- // the frist letter shall NOT be "." or "-"
+ // Cannot start with a '.', '-', or end with a '-'.
//
- if (*addrString == chPeriod ||
- *addrString == chDash )
+ if (*addrString == chPeriod ||
+ *addrString == chDash ||
+ addrString[addrStrLen-1] == chDash)
return false;
- // rightmost domain label starting with digit indicates IP address
- // since top level domain label can only start with an alpha
- // see RFC 2396 Section 3.2.2
- int addrStrLen = XMLString::stringLen(addrString);
+ // rightmost domain label starting with digit indicates IP address
+ // since top level domain label can only start with an alpha
+ // see RFC 2396 Section 3.2.2
+
int lastPeriodPos = XMLString::lastIndexOf(addrString, chPeriod);
// if the string ends with "."
@@ -1369,48 +1456,26 @@ bool XMLUri::isWellFormedAddress(const XMLCh* const addrString)
return false;
}
- if (XMLString::isDigit(addrString[lastPeriodPos + 1]))
+ if (XMLString::isDigit(addrString[lastPeriodPos + 1]))
{
- //
- // IPv4address = 1*digit "." 1*digit "." 1*digit "." 1*digit
- //
- // make sure that
- // 1) we see only digits and dot separators,
- // 2) that any dot separator is preceded and followed by a digit
- // 3) that we find 3 dots
-
- int numDots = 0;
- for (int i = 0; i < addrStrLen; i++)
- {
- if (addrString[i] == chPeriod)
- {
- if (((i > 0) &&
- (!XMLString::isDigit(addrString[i-1]))) ||
- ((i + 1 < addrStrLen) &&
- (!XMLString::isDigit(addrString[i+1]))) )
- {
- return false;
- }
- numDots++;
- }
- else if (!XMLString::isDigit(addrString[i]))
- {
- return false;
- }
- } //for
-
- if (numDots != 3)
- {
- return false;
- }
+ return isWellFormedIPv4Address(addrString, addrStrLen);
} // end of IPv4address
- else
+ else
{
//
// hostname = *( domainlabel "." ) toplabel [ "." ]
// domainlabel = alphanum | alphanum *( alphanum | "-" ) alphanum
// toplabel = alpha | alpha *( alphanum | "-" ) alphanum
- //
+
+ // RFC 2396 states that hostnames take the form described in
+ // RFC 1034 (Section 3) and RFC 1123 (Section 2.1). According
+ // to RFC 1034, hostnames are limited to 255 characters.
+ if (addrStrLen > 255) {
+ return false;
+ }
+
+ unsigned int labelCharCount = 0;
+
// domain labels can contain alphanumerics and '-"
// but must start and end with an alphanumeric
for (int i = 0; i < addrStrLen; i++)
@@ -1424,18 +1489,232 @@ bool XMLUri::isWellFormedAddress(const XMLCh* const addrString)
{
return false;
}
+ labelCharCount = 0;
}
else if (!XMLString::isAlphaNum(addrString[i]) &&
addrString[i] != chDash)
{
return false;
}
+ // RFC 1034: Labels must be 63 characters or less.
+ else if (++labelCharCount > 63) {
+ return false;
+ }
} //for
}
return true;
}
+//
+// RFC 2732 amended RFC 2396 by replacing the definition
+// of IPv4address with the one defined by RFC 2373.
+//
+// IPv4address = 1*3DIGIT "." 1*3DIGIT "." 1*3DIGIT "." 1*3DIGIT
+//
+bool XMLUri::isWellFormedIPv4Address(const XMLCh* const addr, const int& length)
+{
+ int numDots = 0;
+ int numDigits = 0;
+
+ //
+ // IPv4address = 1*3DIGIT "." 1*3DIGIT "." 1*3DIGIT "." 1*3DIGIT
+ //
+ // make sure that
+ // 1) we see only digits and dot separators,
+ // 2) that any dot separator is preceded and followed by a digit
+ // 3) that we find 3 dots
+ // 4) that each segment contains 1 to 3 digits.
+ // 5) that each segment is not greater than 255.
+
+ for (int i = 0; i < length; ++i)
+ {
+ if (addr[i] == chPeriod)
+ {
+ if (((i > 0) &&
+ (!XMLString::isDigit(addr[i-1]))) ||
+ ((i + 1 < length) &&
+ (!XMLString::isDigit(addr[i+1]))) )
+ {
+ return false;
+ }
+ numDigits = 0;
+ if (++numDots > 3)
+ {
+ return false;
+ }
+ }
+ else if (!XMLString::isDigit(addr[i]))
+ {
+ return false;
+ }
+ // Check that that there are no more than three digits
+ // in this segment.
+ else if (++numDigits > 3)
+ {
+ return false;
+ }
+ // Check that this segment is not greater than 255.
+ else if (numDigits == 3)
+ {
+ XMLCh first = addr[i-2];
+ XMLCh second = addr[i-1];
+ XMLCh last = addr[i];
+ if (!(first < chDigit_2 ||
+ (first == chDigit_2 &&
+ (second < chDigit_5 ||
+ (second == chDigit_5 && last <= chDigit_5)))))
+ {
+ return false;
+ }
+ }
+ } //for
+ return (numDots == 3);
+}
+
+//
+// IPv6reference = "[" IPv6address "]"
+//
+bool XMLUri::isWellFormedIPv6Reference(const XMLCh* const addr, const int& length)
+{
+ int index = 1;
+ int end = length-1;
+
+ // Check if string is a potential match for IPv6reference.
+ if (!(length > 2 && addr[0] == chOpenSquare && addr[end] == chCloseSquare))
+ {
+ return false;
+ }
+
+ // Counter for the number of 16-bit sections read in the address.
+ int counter = 0;
+
+ // Scan hex sequence before possible '::' or IPv4 address.
+ index = scanHexSequence(addr, index, end, counter);
+ if (index == -1)
+ {
+ return false;
+ }
+ // Address must contain 128-bits of information.
+ else if (index == end)
+ {
+ return (counter == 8);
+ }
+
+ if (index+1 < end && addr[index] == chColon)
+ {
+ if (addr[index+1] == chColon)
+ {
+ // '::' represents at least one 16-bit group of zeros.
+ if (++counter > 8)
+ {
+ return false;
+ }
+ index += 2;
+ // Trailing zeros will fill out the rest of the address.
+ if (index == end)
+ {
+ return true;
+ }
+ }
+ // If the second character wasn't ':', in order to be valid,
+ // the remainder of the string must match IPv4Address,
+ // and we must have read exactly 6 16-bit groups.
+ else
+ {
+ if (counter == 6)
+ {
+ XMLCh* ipVfour = (XMLCh*)
+ XMLPlatformUtils::fgMemoryManager->allocate((length+1) * sizeof(XMLCh));
+ XMLString::subString(ipVfour, addr, index+1, end);
+ ArrayJanitor<XMLCh> janitor(ipVfour, XMLPlatformUtils::fgMemoryManager);
+ int newLength = XMLString::stringLen(ipVfour);
+ return isWellFormedIPv4Address(ipVfour, newLength);
+ }
+ else
+ {
+ return false;
+ }
+ }
+ }
+ else
+ {
+ return false;
+ }
+
+ // 3. Scan hex sequence after '::'.
+ int prevCount = counter;
+ index = scanHexSequence(addr, index, end, counter);
+ if (index == -1)
+ {
+ return false;
+ }
+ // If this is the end of the address then
+ // we've got 128-bits of information.
+ else if (index == end)
+ {
+ return true;
+ }
+
+ // The address ends in an IPv4 address, or it is invalid.
+ // scanHexSequence has already made sure that we have the right number of bits.
+ XMLCh* ipVfour =
+ (XMLCh*) XMLPlatformUtils::fgMemoryManager->allocate((length+1) * sizeof(XMLCh));
+ XMLString::subString(ipVfour, addr, (counter > prevCount) ? index+1 : index, end);
+ ArrayJanitor<XMLCh> janitor(ipVfour, XMLPlatformUtils::fgMemoryManager);
+ int newLength = XMLString::stringLen(ipVfour);
+ return isWellFormedIPv4Address(ipVfour, newLength);
+}
+
+//
+// For use with isWellFormedIPv6Reference only.
+//
+int XMLUri::scanHexSequence (const XMLCh* const addr, int index, int end, int& counter)
+{
+ XMLCh testChar = chNull;
+ int numDigits = 0;
+ int start = index;
+
+ // Trying to match the following productions:
+ // hexseq = hex4 *( ":" hex4)
+ // hex4 = 1*4HEXDIG
+ for (; index < end; ++index)
+ {
+ testChar = addr[index];
+ if (testChar == chColon)
+ {
+ // IPv6 addresses are 128-bit, so there can be at most eight sections.
+ if (numDigits > 0 && ++counter > 8)
+ {
+ return -1;
+ }
+ // This could be '::'.
+ if (numDigits == 0 || ((index+1 < end) && addr[index+1] == chColon))
+ {
+ return index;
+ }
+ numDigits = 0;
+ }
+ // This might be invalid or an IPv4address. If it's potentially an IPv4address,
+ // backup to just after the last valid character that matches hexseq.
+ else if (!XMLString::isHex(testChar))
+ {
+ if (testChar == chPeriod && numDigits < 4 && numDigits > 0 && counter <= 6)
+ {
+ int back = index - numDigits - 1;
+ return (back >= start) ? back : start;
+ }
+ return -1;
+ }
+ // There can be at most 4 hex digits per group.
+ else if (++numDigits > 4)
+ {
+ return -1;
+ }
+ }
+ return (numDigits > 0 && ++counter <= 8) ? end : -1;
+}
+
bool XMLUri::isGenericURI()
{
return (getHost() != 0);
diff --git a/src/xercesc/util/XMLUri.hpp b/src/xercesc/util/XMLUri.hpp
index 8a1cd1fb64d1828d970232996737d2a210d2f778..8a5dd554fa1a425c1be53b3880746a28223abf74 100644
--- a/src/xercesc/util/XMLUri.hpp
+++ b/src/xercesc/util/XMLUri.hpp
@@ -57,6 +57,32 @@
/*
* $Id$
* $Log$
+ * Revision 1.10 2003/07/25 10:15:16 gareth
+ * Patch by Michael Glavassevich
+ *
+ * The patch fixes Bugzilla #19787, #20006, #20009, #20010 and #20287, and
+ * several other issues. A summary of the changes is listed below:
+ *
+ * 1. Added '[' and ']' to reserved characters as per RFC 2732.
+ * 2. '[' and ']' added in RFC 2732, are not allowed in path segments, but
+ * may appear in the opaque part.
+ * 3. No URI can begin with a ':'.
+ * 4. URI has no scheme if ':' occurs in a URI after '?' or '#', it's part of
+ * the query string or fragment.
+ * 5. Whitespace (even escaped as %20) is not permitted in the authority
+ * portion of a URI.
+ * 6. IPv4 addresses must match 1*3DIGIT "." 1*3DIGIT "." 1*3DIGIT "."
+ * 1*3DIGIT. Since RFC 2732.
+ * 7. IPv4 addresses are 32-bit, therefore no segment may be larger than 255.
+ * This isn't expressed by the grammar.
+ * 8. Hostnames cannot end with a '-'.
+ * 9. Labels in a hostname must be 63 bytes or less [RFC 1034].
+ * 10. Hostnames may be no longer than 255 bytes [RFC 1034]. (That
+ * restriction was already there. I just moved it inwards.
+ * 11. Added support for IPv6 references added in RFC 2732. URIs such as
+ * http://[::ffff:1.2.3.4] are valid. The BNF in RFC 2373 isn't correct. IPv6
+ * addresses are read according to section 2.2 of RFC 2373.
+ *
* Revision 1.9 2003/05/16 06:01:53 knoaman
* Partial implementation of the configurable memory manager.
*
@@ -342,6 +368,7 @@ private:
static const XMLCh MARK_CHARACTERS[];
static const XMLCh SCHEME_CHARACTERS[];
static const XMLCh USERINFO_CHARACTERS[];
+ static const XMLCh PATH_CHARACTERS[];
//helper method for getUriText
void buildFullText();
@@ -356,6 +383,13 @@ private:
* @return true if the string contains any reserved characters
*/
static bool isReservedCharacter(const XMLCh theChar);
+
+ /**
+ * Determine whether a character is a path character:
+ *
+ * @return true if the character is path character
+ */
+ static bool isPathCharacter(const XMLCh theChar);
/**
* Determine whether a char is an unreserved character.
@@ -381,11 +415,14 @@ private:
static void isConformantUserInfo(const XMLCh* const userInfo);
/**
* Determine whether a string is syntactically capable of representing
- * a valid IPv4 address or the domain name of a network host.
+ * a valid IPv4 address, IPv6 reference or the domain name of a network host.
*
* A valid IPv4 address consists of four decimal digit groups
* separated by a '.'.
*
+ * See RFC 2732 Section 3, and RFC 2373 Section 2.2, for the
+ * definition of IPv6 references.
+ *
* A hostname consists of domain labels (each of which must begin and
* end with an alphanumeric but may contain '-') separated by a '.'.
* See RFC 2396 Section 3.2.2.
@@ -394,6 +431,49 @@ private:
* or hostname
*/
static bool isWellFormedAddress(const XMLCh* const addr);
+
+ /**
+ * Determines whether a string is an IPv4 address as defined by
+ * RFC 2373, and under the further constraint that it must be a 32-bit
+ * address. Though not expressed in the grammar, in order to satisfy
+ * the 32-bit address constraint, each segment of the address cannot
+ * be greater than 255 (8 bits of information).
+ *
+ * @return true if the string is a syntactically valid IPv4 address
+ */
+ static bool isWellFormedIPv4Address(const XMLCh* const addr, const int& length);
+
+ /**
+ * Determines whether a string is an IPv6 reference as defined
+ * by RFC 2732, where IPv6address is defined in RFC 2373. The
+ * IPv6 address is parsed according to Section 2.2 of RFC 2373,
+ * with the additional constraint that the address be composed of
+ * 128 bits of information.
+ *
+ * Note: The BNF expressed in RFC 2373 Appendix B does not
+ * accurately describe section 2.2, and was in fact removed from
+ * RFC 3513, the successor of RFC 2373.
+ *
+ * @return true if the string is a syntactically valid IPv6 reference
+ */
+ static bool isWellFormedIPv6Reference(const XMLCh* const addr, const int& length);
+
+ /**
+ * Helper function for isWellFormedIPv6Reference which scans the
+ * hex sequences of an IPv6 address. It returns the index of the
+ * next character to scan in the address, or -1 if the string
+ * cannot match a valid IPv6 address.
+ *
+ * @param address the string to be scanned
+ * @param index the beginning index (inclusive)
+ * @param end the ending index (exclusive)
+ * @param counter a counter for the number of 16-bit sections read
+ * in the address
+ *
+ * @return the index of the next character to scan, or -1 if the
+ * string cannot match a valid IPv6 address
+ */
+ static int scanHexSequence (const XMLCh* const addr, int index, int end, int& counter);
/**
* Get the indicator as to whether this URI uses the "generic URI"
@@ -541,6 +621,11 @@ inline bool XMLUri::isReservedCharacter(const XMLCh theChar)
return (XMLString::indexOf(RESERVED_CHARACTERS, theChar) != -1);
}
+inline bool XMLUri::isPathCharacter(const XMLCh theChar)
+{
+ return (XMLString::indexOf(PATH_CHARACTERS, theChar) != -1);
+}
+
inline bool XMLUri::isUnreservedCharacter(const XMLCh theChar)
{
return (XMLString::isAlphaNum(theChar) ||